| Popis: |
The past years have seen the telecommunications network and services increase in their complexity due to changes in traffic patterns, usage of the network and applications. This is in part driven by the growing customer base, services, broadband usage and complicated by the presence of multiple types of equipment vendors, technology and software systems. One consequence of these is the increase in the breakdown of telecommunication equipment serving end-users. The equipment failure, in turn, necessitates field technicians to visit onsite premise frequently to troubleshoot and resolve customer trouble tickets (CTTs). In this paper, Markov Process (MP) is used to represent the CTTs as a sequence of events and analyze the tendency of one type CTT to be followed by another type of CTT. By observing existing sequences of CTT types, a MP model can approximate the likelihood of the next CTT occurring followed by current CTT type. With this knowledge, technicians can act proactively to preempt network problems. The analysis uses one year of CTT records raised by the subscribers of a leading broadband service. The analysis observes that in most cases current CTTs are influenced by previous CTT types and mean time between failures (MTBF) is different for the same CTT types. |
